Comparative Durability of GFRP bars in Concrete and in Simulated Concrete Environment Abstract Many studies suggest that the durability of glass fiber reinforced polymer (GFRP) bars in simulated concrete pore solution is very different than an actual concrete environment. This study therefore provides a comparative evaluation on the durability of GFRP bars in concrete and in simulated concrete environment through the investigation of their interlaminar shear strength. It focuses on the evaluation of the physical, mechanical and micro-structural properties of GFRP bars under high moisture, saltwater and alkali environments. Bare GFRP bars and cement-embedded GFRP bars were immersed in solutions at different temperatures (23 o C, 60 o C and 80 o C) and exposure times (28 days, 56 days and 112 days). The results showed that the percentage water uptake and the apparent diffusivity of the GFRP bars were strongly dependent on the type of immersion solution and temperature. Direct immersion in solution were found to more severely deteriorate the interlaminar shear strength of bare GFRP bars than cement embedded bars. Moreover, alkaline solution was more aggressive to GFRP bars than tap water and saline solution affecting its fiber and matrix interface, and chemical structure for bars exposed after 112 days. As a result of this study, master curves and time shift factor were developed to correlate the interlaminar shear strength retention from the accelerated aging test to the service life of GFRP bars in actual concrete environment. Keywords: GFRP reinforcing bars; interlaminar shear strength; cement embedded bars; comparative durability; alkaline solution; time shift factor. Introduction Glass fibre-reinforced polymer (GFRP) bars have emerged as a promising and cost-effective replacement for steel to increase the useful life of reinforced concrete structures exposed in severely aggressive environments. Research and development related to this high tensile strength, lightweight, non-corrosive, non-magnetic, and non-electrical conductive reinforcing material has been carried out extensively in the US, Canada, Europe, and Japan (Benmokrane et al. 2017) leading to the many successful field applications of GFRP reinforced concrete structures including highway bridges and barriers, pavements and parking garages, storage facilities for chemical and wastewater treatment plants, magnetic-resonance-imaging facilities, detector loops in railway lines, and temporary structures such as soft-eyes in underground excavations and tunnelling works. This composite reinforcing material is now also being increasingly utilised in Australia as a main reinforcement in concrete structures that operate near coastline and in aggressive soils (Maranan et al. 2016, Manalo et al. 2014, Ferdous et al. 2015). As internal reinforcement, GFRP bars are continuously subjected to alkaline environment due to its surrounding concrete and other environmental conditions that may affect their physical, mechanical and long-term durability properties. Many researchers have suggested that the high alkalinity of concrete pore-solution is an aggressive environment for the GFRP bars, which could cause damages at the glass fibers and/or deteriorate the fiber-resin interface (Chin et al. 2001, Tannous and Saadatmanesh 1999, Karbhari and Zhang 2003, Belarbi and Wang 2012, Benmokrane et al. 2017a). Similarly, infrastructure systems are exposed to external agents during their service life including high moisture, alkalinity and
Transcript
Comparative Durability of GFRP bars in Concrete and in Simulated Concrete
Environment
Abstract
Many studies suggest that the durability of glass fiber reinforced polymer (GFRP) bars in
simulated concrete pore solution is very different than an actual concrete environment. This
study therefore provides a comparative evaluation on the durability of GFRP bars in concrete
and in simulated concrete environment through the investigation of their interlaminar shear
strength. It focuses on the evaluation of the physical, mechanical and micro-structural
properties of GFRP bars under high moisture, saltwater and alkali environments. Bare GFRP
bars and cement-embedded GFRP bars were immersed in solutions at different temperatures
(23oC, 60oC and 80oC) and exposure times (28 days, 56 days and 112 days). The results showed
that the percentage water uptake and the apparent diffusivity of the GFRP bars were strongly
dependent on the type of immersion solution and temperature. Direct immersion in solution
were found to more severely deteriorate the interlaminar shear strength of bare GFRP bars than
cement embedded bars. Moreover, alkaline solution was more aggressive to GFRP bars than
tap water and saline solution affecting its fiber and matrix interface, and chemical structure for
bars exposed after 112 days. As a result of this study, master curves and time shift factor were
developed to correlate the interlaminar shear strength retention from the accelerated aging test
to the service life of GFRP bars in actual concrete environment.
Water absorption at 24 hrs (%) ASTM D570-98 (2010) 15 0.15 0.01
Water absorption at saturation
(%)
ASTM D570-98 (2010) 15 0.19 0.01
Cure ratio (%) ASTM E 1356-08 (2014) 15 100 0
Tg (°C) ASTM E 1356-08 (2014) 15 125.8 1.3
Mechanical properties
Flexural strength, fu (MPa) ASTM D4476/D4476M-14
(2014)
6 1623.7 58.2
Interlaminar shear strength, Su
(MPa)
ASTM D4475-02 (2016) 6 54.7 1.1
Tensile strength, ft (MPa) ASTM D7205/D7205M-06
(2011)
6 1315.3 31.1
Tensile modulus, E (GPa) 62.5 0.4
Tensile strain, e 2.3 0.1
Test specimens
A total of 324 specimens were prepared, conditioned and tested. The bars were divided into
two series: (1) GFRP bars and (2) the cement-embedded GFRP bars as shown in Figure 1. The
test specimens were cut into 40 mm length (approximately four times bar diameter, 4db) from
the GFRP bars manufactured from the same production lot. The cement-embedded GFRP bars
(Figure 1b) were placed centrally into a 60 mm long PVC pipe filled with cement paste of 25
mm diameter cylindrical shape providing a cover of 7.5 mm around the bar and 10 mm on both
ends. The GFRP bars embedded in cement are prepared to simulate the environmental
conditions of reinforcements inside concrete structures. However, the thickness of the cement
cover was kept to a minimum for easy removal of the bars after conditioning. The cement paste
has a water-to-cement ratio of 0.45. Six (6) replicates were prepared for each specimen type as
summarized in Table 1. The pH of the hardened cement paste was 12.8 as measured according
to ASTM D4972 (2013).
a. GFRP bars b. Cement-embedded GFRP bars
Figure 1: Specimen types
Table 1 - Specimen details.
Tap Water
Exposure time, days
Subtotal Total 28 56 112
Bar
Only
Bar in
Cement
Bar
Only
Bar in
Cement
Bar
Only
Bar in
Cement
Temp, oC
RT 6 6 6 6 6 6 36
108 60 6 6 6 6 6 6 36
80 6 6 6 6 6 6 36
Alkaline Solution
Exposure time, days
Subtotal Total 28 56 112
Bar
Only
Bar in
Cement
Bar
Only
Bar in
Cement
Bar
Only
Bar in
Cement
RT 6 6 6 6 6 6 36 108
Temp, oC
60 6 6 6 6 6 6 36
80 6 6 6 6 6 6 36
Saline Solution
Exposure time, days
Subtotal Total 28 56 112
Bar
Only
Bar in
Cement
Bar
Only
Bar in
Cement
Bar
Only
Bar in
Cement
Temp, oC
RT 6 6 6 6 6 6 36
108 60 6 6 6 6 6 6 36
80 6 6 6 6 6 6 36
Conditioning
The conditioning of the specimens was done by immersing the GFRP bars and cement-
embedded GFRP bars in three different types of solutions, i.e. tap water (TW), saline solution
(SS) and alkaline solution (AS) for 28 days, 56 days, and 112 days (672 hours, 1344 hours and
2688 hours, respectively) under three different temperatures (23°C, 60°C, and 80°C) as shown
in Figure 2. These conditioning temperatures were selected as they are well below the glass
transition temperature of the GFRP bars of 125.8oC as reported in Table 1. Bank et al. (2003)
suggested these levels of temperature will accelerate the degradation effect of aging but ensure
that the GFRP bars will not be subjected to a change of degradation mechanism. The tap water
simulates high moisture condition while the saline solution was prepared in accordance with
ASTM D 1141-98 (2013) using 3.5% NaCl by weight solution to match the salinity of saltwater
in a marine environment. On the other hand, the alkaline solution had a pH of 12.7 to simulate
the concrete environment and prepared in accordance with ASTM D7705 (2012). During the
conditioning, the glass containers were constantly checked and refilled as needed to make sure
that the specimens were completely submerged and covered to minimise evaporation of the
liquid solution. The conditioning elevated temperatures of 60°C and 80°C were achieved by
placing the glass containers inside a programmable oven.
Figure 2: Conditioning of GFRP bars
Moisture Uptake
The percentage moisture uptake of the conditioned GFRP bars (only those samples which are
not embedded in cement) were measured periodically following ASTM D570 (2010). The
samples were weighed to determine the initial mass (m0) prior to conditioning. After immersion
for each exposure duration, the specimen was removed from the solution, quickly washed with
tap water, dried using tissue paper, and then weighed immediately (m1B). Lastly, the specimen
was post-conditioned, and then reweighed (m1A). The percentage moisture uptake was
expressed as the ratio between the weight of the water penetrated through the bars (weight of
wet specimen – weight of dry specimen) and the weight of the dry bars, and was calculated
using Equation 1:
1 0
0
(%) 100Am mMass gain
m
Equation 1
Short-Beam Shear Test
The long-term performance of the GFRP bars was evaluated through investigation of their
interlaminar shear strength (ILSS) in accordance with ASTM D4475-02 (2016). For cement-
embedded GFRP bars, the bars were carefully extracted to avoid damage using a small hammer
and a flat head. It is important to note that there was no crack in the cement cover after
conditioning indicating that there was no expansion in the bars and the solution did not
penetrate through cracks. Short-beam test of GFRP bars with a clear span of 3db and
overhanging length of 0.5db was performed. Micelli and Nanni (2004) indicated that this span
will avoid flexural effects that could change the desired shear failure mode of the bars. Six
samples from each group was tested to examine the inter-laminar shear strength of the
conditioned and unconditioned GFRP bars. The inter-laminar shear testing was performed
using MTS 100kN testing machine with a displacement-control rate of 1.3 mm/min. The actual
test set-up is shown in Figure 3a. The inter-laminar shear strength, Su, was calculated as: Su =
0.849P/d2, where P is the shear failure load (N), and d is the bar diameter (mm). The average
and standard deviation of the ILSS for all the tested specimens are presented in Table 4.
The specimens was designated as based on the cement-embedded (CE) or not, followed by the
type of conditioning environment (TW for tap water, SS for saline solution, and AS for alkaline
solution), followed by the level of temperature (RT for room temperature or 23oC, 60 for 60oC,
and 80 for 80oC), and finally the exposure duration (28, 56, and 112 days). For example,
specimen CE-TW-60-112 represents the cement embedded GFRP bars conditioned in tap water
at a temperature of 60oC for 112 days.
a. Test set-up b. Failure of control and conditioned bars
Figure 3: Short-Beam Shear Test
Results and Discussion
Moisture Uptake
The moisture uptake is an important factor that could significantly affect the mechanical and
durability properties of GFRP bars (Benmokrane et al. 2001). Micelli and Nanni (2004)
indicated that there is a strong evidence that the rate of degradation of GFRP bars exposed to
fluid environment is related to the rate of sorption of the fluid. GFRP bars embedded in
reinforced concrete elements can absorb moisture and water, which penetrate through the resin
affecting the fiber-resin interface (Bradley and Grant 1994, Kawagoe et al. 2001). Moreover,
the volume expansion of the water at low temperature causes degradation in the polymeric
resin, thus, reduction in shear strengths of GFRP bars (Robert and Benmokrane 2010).
In this study, the percentage moisture uptake at different conditioning times is used to obtain
quantitative information on the diffusion properties of GFRP bars in different solutions. Figure
4 shows the plot of the moisture uptake at saturation of bare GFRP bars conditioned in tap
water (TW), alkaline solution (AS), and saline solution (SS) and in different temperatures
(23°C, 60°C, and 80°C) against the square root of the exposure time in seconds. On the other
hand, Table 2 summarises the peak moisture uptake of the GFRP bars (Mm) and the apparent
diffusivity (D) in the different solution and conditioning temperature as calculated using the
Fick’s equation (Equation 2) for a bar-shaped specimen presented by Aiello et al. (2006):
𝐷 = 𝜋 (𝑑
4.𝑀𝑚)
2(
𝑀2−𝑀1
√𝑡2− √𝑡1)
2 Equation 2
where d is the nominal diameter of the bar, M1 and M2 are the moisture contents at times t1 and
t2, respectively. Both at times t1 and t2, respectively. Both t1 and t2 are taken at times where the
change in moisture uptake is relatively linearly varying with the square root of time as
suggested by Karbhari and Zhang (2003).
There is obvious amount of scatter in the measured percentage moisture uptake primarily due
to the relatively small mass change. In general, the moisture uptake process of GFRP bars
followed the Fick’s Second Law of Diffusion. The moisture uptake increased linearly in the
first 28 days (s1/2 of around 1500) and tend to stabilise after that with the increasing square root
of exposure time due to the bars already reached saturation. This two-stage moisture uptake is
a typical behaviour of high volume fraction unidirectional composites as also suggested by
Karbhari and Xian (2009). Of particular note is the difference in percentage moisture uptake
of the GFRP bars depending on the type of solution and temperature, with the percentage
moisture uptake higher for higher temperature exposure indicating that temperature has a direct
effect on the water absorption. As with the percentage moisture uptake, apparent diffusivity
values also increase with the temperature. For similar degree of temperature exposure, the
GFRP bars conditioned in the alkaline solution has the highest moisture uptake (0.16%-0.22%)
as shown in Fig. 4c followed by the bars conditioned in tap water (0.14%-0.18%) and saline
solution (0.12%-0.15%) as shown in Figs. 4a and 4b, respectively. The high moisture uptake
in AS is due to the free OH-ions in the solution which can break the Si-O-Si structures and
promote diffusion of more water inside the bar as found by Chen et al. (2007). On the other
hand, Bank et al. (2003) indicated that the higher moisture uptake of composites in water than
in saline solution can be attributed to the larger size of salt ions compared to that of water.
Overall, it is good to note that the moisture uptake of the bars is very low and is similar to that
of the percentage absorption of the straight GFRP bars samples (0.14%) measured by
Benmokrane et al. (2018). Moreover, the measured apparent diffusivity which ranges from
0.71 x 10-7 mm2/s compares well with the values reported by Karbhari et al. (2002) for E-
glass/vinylester composites exposed to concrete based alkali solutions. The low moisture
uptake and apparent diffusivity values were also due to the less hydrophilic properties of the
vinylester resin as it contain less polar ester moieties as found by Chin et al. (2001). This
behaviour also suggests that the moisture was absorbed only through the resin rich surface of
the GFRP bars as suggested by Mouzakis et al. (2008). The effect of this absorbed solution to
the fiber-matrix interface was evaluated through the interlaminar shear strength test and
discussed in the next section.
a. TW b. SW
c. AS
Figure 4: Moisture absorption process of GFRP bars conditioned at different solutions and
temperatures.
Table 2 – Comparison of the peak moisture absorption (Mm) and apparent diffusivity (D)
Temperature
(°C)
Tap Water Saline Solution Alkaline Solution
Mm (%) D
(×10-7 mm2/s)
Mm (%) D
(×10-7 mm2/s)
Mm (%) D
(×10-7 mm2/s)
23 0.14 1.01 0.12 0.71 0.16 1.75
60 0.15 1.67 0.14 1.32 0.17 2.13
80 0.18 2.55 0.15 1.92 0.22 3.14
0.00%
0.05%
0.10%
0.15%
0.20%
0.25%
0 1000 2000 3000 4000
Wat
er A
bso
rpti
on
Square Root of Time (s1/2)
23°C60°C80°C
0.00%
0.05%
0.10%
0.15%
0.20%
0.25%
0 1000 2000 3000 4000
Wat
er A
bso
rpti
on
Square Root of Time (s1/2)
23°C60°C80°C
0.00%
0.05%
0.10%
0.15%
0.20%
0.25%
0 1000 2000 3000 4000
Wat
er A
bso
rpti
on
Square Root of Time (s1/2)
23°C60°C80°C
Interlaminar Shear Strength (ILSS)
The long-term properties of the bare and cement -embedded GFRP bars in different solutions
was evaluated through investigation of their interlaminar shear strength (ILSS). Gooranorimi
and Nanni (2017) and Benmokrane et al. (2018) used this property as a useful parameter for
durability evaluation especially for GFRP bars in in-field conditions as the bars that can be
extracted from the concrete core mostly has limited lengths.
All specimens failed because of the horizontal shear originated at the edge of the bars and
developed along the length as shown in Figure 3b. A minor difference observed is the location
of the inter-laminar shear crack along the bar’s cross-section. In the control and conditioned
GFRP bars after 28 and 56 days, the crack mostly occurred at the mid-depth of the bars. On the
other hand, the crack in the bars conditioned after 112 days occurred below the mid-depth as
shown in Figure 3b (right) indicating a weaker fiber–matrix interface in this location. This
observed crack at the lower portion of the conditioned bars can be due to the bars already
saturated with solution at the outermost resin rich surface, which was degraded and created a
weak zones along which the transverse cracking could easily take place during testing.
Tables 3 and 4 summarise the average and standard deviation of the ILSS of bare and cement-
embedded GFRP bars exposed to different solutions and temperatures, respectively. The low
values of the standard deviation (maximum of 8.3%) suggest the good extraction of the bars,
as Gooranorimi and Nanni (2017) and Benmokrane et al. (2018) indicated that this procedure
is a major factor that may influence the ILSS of GFRP bars. The average ILSS of the
unconditioned bars was 54.9 MPa with a standard deviation of 0.7 MPa. This is comparable to
the properties reported by Benmokrane et al. (2017) in Table 1 as these bars were obtained
from the same production lot.
Table 3 - Interlaminar shear strength in MPa of bare GFRP bars.
Temp
(°C)
Tap Water Saline Solution Alkaline Solution
28
days
56
days
112
days
28
days
56
days
112
days
28
days
56
days
112
days
23 47.4
(0.1)
37.2
(0.7)
24.3
(0.7)
51.9
(0.8)
44.6
(0.5)
37.6
(0.4)
44.6
(0.2)
32.1
(0.2)
23.2
(0.8)
60 41.7
(0.3)
31.5
(0.7)
18.4
(1.2)
49.3
(1.6)
39.1
(0.4)
29.8
(0.6)
34.5
(0.5)
27.4
(0.4)
16.6
(0.7)
80 37.4
(0.7)
26.6
(0.6)
9.3
(0.8)
46.6
(0.3)
35.3
(0.7)
23.7
(0.6)
28.7
(2.4)
20.9
(0.9)
12.9
(0.9)
The numbers inside ( ) are the standard deviation
Table 4 – Interlaminar shear strength in MPa of cement-embedded GFRP bars.
Temp
(°C)
Tap Water Saline Solution Alkaline Solution
28
days
56
days
112
days
28
days
56
days
112
days
28
days
56
days
112
days
23 50.3
(2.0)
49.3
(0.8)
48.1
(0.3)
52.5
(1.3)
50.8
(0.8)
49.7
(1.9)
47.5
(0.2)
45.9
(0.4)
44.1
(0.5)
60 48.8
(0.6)
47.0
(0.6)
45.6
(0.9)
50.8
(1.0)
49.4
(0.5)
47.7
(0.7)
43.4
(0.2)
41.5
(0.7)
40.3
(1.0)
80 43.5
(0.6)
41.5
(0.3)
38.5
(0.4)
47.3
(0.9)
46.1
(0.2)
44.6
(0.2)
41.3
(1.8)
39.1
(0.8)
36.3
(0.4) The numbers inside ( ) are the standard deviation
Generally, the ILSS of the GFRP bars reported in Tables 3 and 4 decreases as the exposure
temperature and duration increases in all exposure conditions. This can be attributed to the
increase in percentage moisture uptake of the bars with exposure duration which leads to
degradation of the fibre matrix interface, resulting in the decrease in the ILSS. When the bars
absorbed moisture, the resin rich surface degrades and the bond between the matrix and fibers
located at the outer part of the bar will gradually reduce and the bar resistance will start to
decrease. While the water absorption of the GFRP bars was low (maximum of 0.22% for AS-
conditioned bars after 112 days) indicating that only a very thin layer was affected by the
solution, the softening of the resin at the edge of the bars was very critical under short-beam
shear test as this location is subjected to high level of shear stress. Nkurunziza et al. (2005)
highlighted that the interface between glass fibers and the resin controls the resistance of GFRP
bars to different environments. However, the ILSS of cement-embedded GFRP bars were
generally higher than that of the bare GFRP bars for similar immersion conditions. This
behaviour can be attributed to the limited availability of the moisture around the bars and the
lower temperature condition for the cement-embedded bars than the bare bars. Ceroni et al.
(2006) indicated that the cut ends expose directly the fibers to external environment giving
undesirable effects in the durability of the bars. The increased in temperature further increased
the water absorption, which resulted in a faster degradation of the fiber and matrix interface
leading to the decrease in the ILSS of the GFRP bars.
ILSS retention
The durability performance of the GFRP bars was appraised based from the ILSS retention
after conditioning to different solutions and exposure durations. The ILSS retention was
calculated by dividing the average ILSS shear strength for each bars reported in Tables 3 and
4 to that of the average property value for the reference specimen. Figure 5 shows the ILSS
retention values of the conditioned bare GFRP bars while Figure 6 shows the ILSS retention
values of the cement-embedded GFRP bars. From Figure 5, the following important
information were observed:
The AS is the most aggressive solution to GFRP bars affecting its ILSS. This is due to
the highest absorption of the bars in this exposure condition. In the case of immersion
in TW, the ILSS strength retention of bare GFRP bars at the end of 28 days immersion
were 86%, 76% and 68% at 23, 60, and 80°C, respectively. For the same time of
immersion, the ILSS retention were 94%, 90%, and 85%, respectively, for the bars
conditioned in SS and were 81%, 63%, and 52%, respectively, for those conditioned
in AS.
At the end of 56-day immersion, the retention levels of the specimens immersed in tap
water were 68%, 57%, and 48% for the 23, 60, and 80 °C cases, respectively, where as
those immersed in saline solution yielded retention levels of 81%, 71%, and 64%,
respectively. Retention levels of 58%, 47%, and 38%, respectively, were recorded for
alkaline-conditioned specimens.
At the end of 112-day immersion, the retention levels of the specimens immersed in tap
water were 48%, 41%, and 29% for the 23, 60, and 80°C cases, respectively, where as
those immersed in saline solution yielded retention levels of 69%, 54%, and 43%,
respectively. Retention levels of 38%, 30%, and 23%, respectively, were recorded for
alkali-conditioned specimens.
For cement-embedded GFRP bars, the following observations from Figure 6 suggest the
following:
The ILSS strength retention of unembedded GFRP bars at the end of 28 days immersion
in TW were 92%, 89%, and 79% at 23°C, 60°C, and 80°C, respectively. For the
duration of immersion, the ILSS retention were 96%, 93% and 86%, respectively, for
the specimens conditioned in SS and were 87%, 79%, and 75%, respectively, for those
conditioned in AS.
At the end of 56-day immersion, the retention levels of the GFRP bars immersed in TW
were 90%, 86%, and 76% for the 23, 60, and 80°C cases, respectively, where as those
immersed in SS yielded retention levels of 93%, 90%, and 84%, respectively. Retention
levels of 84%, 76%, and 71%, respectively, were recorded for AS-conditioned bars.
At the end of 112-day immersion, the retention levels of the bars immersed in TW were
88%, 83%, and 70% for the 23, 60, and 80°C cases, respectively, where as those
immersed in SS yielded retention levels of 91%, 87%, and 81%, respectively. Retention
levels of 80%, 73%, and 68%, respectively, were recorded for AS-conditioned
specimens. These results clearly that direct immersion in alkali solutions deteriorates
the GFRP bars more severely than in the case of exposure to concrete alkali.
a. 28 days b. 56 days
c. 112 days
Figure 5: Interlaminar shear strength retention of bare GFRP bars
a. 28 days b. 56 days c. 112 days
Figure 6: Interlaminar shear strength retention of cement-embedded GFRP bars
The above experimental results clearly indicated the immediate impact of high temperatures
on the degradation of ILSS of the GFRP bars. It was also observed that the ILSS retention of
the GFRP bars decreases as the exposure temperature and duration increases in all exposure
conditions. The significant degradation in ILSS could be explained by the degradation of the
matrix at higher temperature which in turn could have affected the interface between the fibre
and the matrix. The water molecules penetrated through the fiber-resin interface have a
plasticization effect and hydrolysis reactions causing interfacial fracture. Moreover, the AS is
found the most aggressive to GFRP bars affecting its ILSS. Chen et al. (2007) suggested that
0%
20%
40%
60%
80%
100%
23 60 80
ILS
S R
ete
ntio
n
Temperature, C
TW SS AS
0%
20%
40%
60%
80%
100%
23 60 80
ILS
S R
ete
ntio
n
Temperature, C
TW SS AS
0%
20%
40%
60%
80%
100%
23 60 80
ILS
S R
ete
ntio
n
Temperature, C
TW AS AS
0%
20%
40%
60%
80%
100%
23 60 80
ILS
S R
ete
ntio
n
Temperature, C
TW SS AS
0%
20%
40%
60%
80%
100%
23 60 80
ILS
S R
ete
ntio
n
Temperature, C
TW SS AS
0%
20%
40%
60%
80%
100%
23 60 80
ILS
S R
ete
ntio
n
Temperature, C
TW SS AS
the bond in the vinyl ester matrix is highly prone to the degradation at the presence of the free
OH-ions in the alkaline solution by hydrolysis reactions due to the ester group leading to a
leaching in the matrix and damaging in the glass fibers decreasing the interfacial bond strength.
Furthermore, Micelli and Nanni (2004) indicated that the high pH of the pore water solution
during the hydration of the concrete may cause the chemical attack of the glass fibers. This
agrees with the findings of Nkurunziza et al. (2005) wherein they indicated that the most critical
degradation of GFRP bars takes place in alkaline solution. Similar finding was observed by
Wang et al. (2017) wherein they measured higher rate of degradation in the ILSS to solution
with higher alkali-ion solution. However, the measured reduction in ILSS is significantly lower
compared with the almost 90% reduction in ILSS in GFRP bars immersed in alkaline solution
for 42days at 60oC as measured by Micelli and Nanni (2004). On the contrary, Ashrafi et al.
(2018) measured a maximum reduction of 15% in the ILSS of E-glass FRP bars subjected to
water vapour condensation after 3000 hours as this conditioning only affected the very thin
matrix rich surface of the bars. Similarly, Benmokrane et al. (2017) measured only a 12%
decrease in ILSS for a 9.5 mm diameter GFRP bar after conditioning in an alkaline solution
for 90 days at 60oC.
The decrease in the ILSS retention of GFRP bars exposed to TW and SS are in agreement with
the evaluation conducted by D’Antino et al. (2018) that vinylester-based GFRP bars embedded
within concrete and conditioned in tap water and saline solution at 50-60oC showed significant
strength decrease with increasing exposure time. Wang et al. (2017) measured only a 50%
retention in the interlaminar shear strength of GFRP bars exposed in seawater solution at 55oC
for 84 days. While researchers (Robert et al. 2009; Robert and Benmokrane 2013) have
indicated that tapwater and saltwater have similar effects on the durability of the GFRP bars,
the results of this study suggest a higher decrease in ILSS exposed in tapwater than in seawater.
D’Antino et al. (2018) have a similar observation wherein GFRP bars conditioned in salt
solution were less degraded than those conditioned with plain water at the same temperature.
This can be due to the higher moisture uptake of the GFRP bars in TW than in SS as shown in
Figure 3. This can be also explained by the large salt molecules in the saline solution slowing
the diffusion of water by blocking the paths which water diffuses into the bars as was also
found by Mouzakis et al. (2008). Almusallam et al. (2012) further indicated that the lower
water absorption of GFRP bars in SS than TW can be attributed to the formation of very thin
layer of salt on the specimen surface, especially at higher temperature which decreases the
diffusion rate of the solution into the GFRP bars. In all exposure temperatures considered in
this study, the rate of diffusion of GFRP bars exposed in SS was lower than that of TW. It is
also worth noting that the works of Robert et al. (2009) and Robert and Benmokrane (2013)
considered the tensile strength as an indicator of degradation. This property is governed more
by the properties of the fibres and not the matrix, which can be better evaluated by ILSS test.
For cement-embedded GFRP bars, it can be noticed that the reduction in ILSS increased
gradually with time. In contrast, a significant decrease even only after 28 days can be observed
for bare GFRP bars especially for bars exposed in AS. For example, the bare GFRP bars
retained only 63% of its ILSS while cement embedded bars retained 79% at 60oC exposure in
AS. The 21% decrease in ILSS for cement embedded bars is almost equal to the 23% decrease
in the ILSS measured by Gooranorimi and Nanni (2017) in 15.9 mm diameter GFRP bars
extracted from a concrete bridge deck after 15 years of service. However, it is important to note
that only one extracted sample was tested by these researchers and insufficient to prove the
reliability of the test results. Similarly, Micelli and Nanni (2004) estimated that the alkaline
exposure of GFRP bars for 21 days at 60oC would correspond respectively to 14 years in a real
concrete structure. The higher ILSS retention of cement embedded GFRP bars was due to the
cement which may not be completely saturated, and the moisture from outside of the concrete
is not in direct contact with the bars. The cement cover minimises the water absorption of the
GFRP bars whereas the bare GFRP bars are completely immersed in solution that wet all the
bar external surface. As a result, the reaction of decomposition in the GFRP bars embedded in
the cement would be slower because of the absence of oxidation reaction of the polymer matrix.
SEM and FTIR Observations
Scanning-electron-microscopy (SEM) observations were performed to assess the
microstructure of the GFRP bars before and after conditioning after 112 days using the JEOL
JSM-840A SEM (JEOL, Akishima, Tokyo, Japan). All of the specimens observed under SEM
were cut, polished, and coated with a thin layer of gold–palladium using a vapor-deposit
process. SEM observations were performed in both the cross-section of the bars and at the
fibre-matrix interface. Similarly, Fourier Transformed Infrared Spectroscopy (FTIR) was
conducted to study the changes in the chemical composition of the matrix at the bar surface.
These observations were implemented to determine the potential degradation of the polymer
matrix, glass fibers, or interface, as applicable, due to the penetration of the solution. The aim
was to link these observations to the possible evolution of ILSS and chemical composition of
the bars after conditioning.
SEM
Figure 7 shows the SEM observation at the cross section of the bars at 2500 times
magnification. There was no visible difference observed between the bars and the cement
embedded bars and exposed in different solutions. This can be due to the bar not under the
stress so the fibers and matrix interface is still intact. Moreover, no pores, air bubbles were
observed indicating the high quality of the manufacturing process.
SEM was also performed on the fracture zones near the ends of the GFRP bars conditioned for
112 days after the short beam testing to investigate the mechanisms of failure at the fiber-matrix
interface as shown in Figure 8. The SEM shows that the fracture of the surfaces are dominated
by matrix fracture as the matrix layer at the surface that covers and protects the glass fibers was
lost in some areas. In all specimens, no fibre damage in the internal section of the bar was
observed. As compared to the bare GFRP bars exposed to TW and SS (Figures 8a and 8b,
respectively), which shows resin adhering to fibers, i.e. a good interface, the resin left on the
fibre surface of the bars exposed to AS became less indicating decreased bonding strength
between fibre and resin. Figure 8c shows some of the fibres are smooth and almost no resin
residues, which indicates that the rupture occurred partly at the interface. In contrast, the fiber
surface of the cement embedded GFRP bars had more resin coverage than those bare GFRP
bars but still could have lost a certain adhesion. This observation explains the lower ILSS
retention of the bare GFRP bars compared to that of the cement embedded bars after
conditioning. Moreover, a lot of residual resin covers on the fibre surface was observed for
cement embedded bars exposed to TW and AS (Figures 8d and 8e, respectively), which
suggests a better bonding between the fibre and vinyl-ester resin than the bars exposed to AS
(Figure 8f). This observation also suggests that the integrity of the fiber and matrix interface
can be better evaluated by observing the fracture surface than at the ends of the GFRP bars.
The observed damage of the fibre-resin interface and decrease in the ILSS was due to the
penetration of the fluid that resulted in a moderate moisture content after immersion. The
moisture absorbed by the bars, combined with the temperature of exposure, induces stress in
the material with consequent damage at their interface decreasing the ILSS strength with time.
Mouzakis et al. (2008) indicated that absorbed water can disrupt the interfacial bonds between
the fibre and the matrix. Moreover, Ceroni et al. (2006) indicated that a deterioration of this
interface reduces the capacity of load transfer between fibers resulting in the decrease in the
mechanical properties. Davalos et al. (2011) highlighted that the fibrer-matrix interface
integrity is critical for load transfer between fibers, and the interface degradation weakens the
composite materials. Moreover, Nkurunziza et al. (2005) indicated that the chemical bond
between the coupling agent and the surface of the glass fibres is not stable in the presence of
moisture and alkalis. When the moisture and the alkalis was absorbed by the bars, this bond is
destroyed gradually causing damage to the interface, and reducing the stress transfer efficiency
between the fibers and matrix within the composite. This presence of water combined with a
high pH levels considerably affects the physical and chemical degradation at the fibre-matrix
interface.
(a) TW (b) SS (c) AS
Figure 7 : SEM observations at the cross section of the bars
(a) TW (bare GFRP bars) (b) SS (bare GFRP bars)
(c) AS (bare GFRP bars) (d) TW (cement embedded)
(e) SS (cement embedded) (e) AS (cement embedded)
Figure 8: SEM observations at the failure surface of conditioned GFRP bars
FTIR
The FTIR spectra of the GFRP bars recorded using a Nicolet Magma 550 spectrometer is
shown in Figure 9. In both the bare GFRP bars (Figure 9a) and cement embedded bars (Figure
9b), the FTIR spectra was focused on OH units around 3500 cm-1 and CH groups (around 2900
cm-1). When there is a degradation by hydrolysis, the OH peak dramatically increase as
compare to the CH peak which remains constant. The hydroxyl peak did not show any
significant changes which indicates that no significant hydrolysis of GFRP bars occurred
except for the relatively higher intensity of the O-H stretching band at 3400 cm-1 for the bare
GFRP bars exposed in AS. Chin et al. (2001) suggested that this spectral change is consistent
with ester hydrolysis, in which ester functional groups are converted to hydroxyl and
carboxylic acid products. However, this higher intensity was only observed at the bar surface
indicating that the water absorption was only concentrated in the thin resin rich area of the
GFRP bars. This indicates that ageing might have occurred on the surface of the GFRP bars.
However, it is also important to note that the O-H stretching may not also be due to hydrolysis.
Vinyl ester naturally contains OH, and if non-evaporated water or alkalis are present inside the
bars, the amount of OH will increase too. This degradation in the materials explained the
fracture fibre and matrix interface observed under the SEM and the significant loss in the ILSS
of the bare GFRP bars exposed in the alkaline solution.
(a) GFRP bars (b) Cement embedded bars
Figure 9 : FTIR of GFRP bars
Prediction of long-term behaviour and service life for ILSS of GFRP bars
Aiello et al. (2006) indicated that a reliable prediction of a long-term behaviour of civil
infrastructure upon the action of environmental factors is a complex problem. Similarly, Wang
et al. (2016) highlighted that the durability of fiber-reinforced polymers under different
environments is difficult to ascertain because of non-standardization of various conditioning
effects and variation in material constituent. This requires accelerated aging through
hygrothermal exposure for the long-term assessment of materials durability and relies on the
superposition of temperature and moisture to enhance and speed up environmental degradation.
Moreover, Davalos et al. (2011) highlighted that only a few studies were directed to the
development of life-cycle durability prediction models for FRP bars in concrete environment.
Naya et al. (2013) indicated that the most accurate method is the Arrhenius method for
materials exposed to temperature less than its glass transition temperature. Thus, the long-term
interlaminar shear strength performance of GFRP bars investigated in this study was predicted
in accordance with Arrhenius relation and following the procedure implemented by Bank et al.
(2003). As a requirement, at least three elevated temperatures and three exposure durations are
necessary to perform an accurate prediction based on Arrhenius law (Robert and Benmokrane
2013), which was conducted in this study.
Arrhenius model
Arrhenius equation is used to express the degradation rate for materials with time as denoted
by Nelson (2009) using Equation 3:
𝑘 = 𝐴 × 𝑒(−𝐸𝑎𝑅𝑇
) Equation 3
where 𝑘 is the degradation rate (1/𝑡𝑖𝑚𝑒), 𝐴 is a constant based on material properties, 𝐸𝑎 is
the activation energy, 𝑅 is the universal gas constant, and 𝑇 is the temperature in Kelvin. The
basic assumption in this prediction model is that the material properties of the GFRP bars are
not affected by the temperature during exposure and the rate of degradation is accelerated with
the increase in temperature. Therefore, analysing procedures to identify the factors in the rate
of degradation consist of transforming Eq. (3) to produce a linear equation between the time
4000 3500 3000 2550Wavenumber [cm-1]
0.6
0.4
0.2
0
AbsAS
TW
SW
Control
4000 3500 3000 2550Wavenumber [cm-1]
1
0.8
0.6
0.4
0.2
0
Abs
AS
TW
SW
Control
(𝑡 = 1/𝑘) and the inverse of temperature (1/𝑇) by taking the natural logarithm for the two
sides of the equation, as shown in Equation 4:
ln(1
𝑘) =
−𝐸𝑎
𝑅(
1
𝑇) + ln(𝐴) Equation 4
Afterwards, the Arrhenius method is carried out by plotting the natural logarithm of time
needed to reach 90%, 80%, 70%, and 60% strength retention (𝑆𝑅) of the ILSS of the GFRP
bars with the inverse temperature (1000/𝑇) in Kelvin to obtain the regression coefficient (𝐸𝑎
𝑅)
value. This value was expressed by the slope of the linear equations as can be seen in Figure
10 for bare bars and Figure 11 for embedded bars in concrete, immersed in different liquid
solutions. Regression analyses were done to determine the line-of-best-fit as shown in these
figures. The straight lines were nearly parallel to each other, indicating that the accelerated
aging tests were valid, and this model may be applied to describe the ILSS degradation of
GFRP bars. Moreover, the R2 of the regression line is at least 0.95 which is well above 0.80 as
indicated by Benmokrane et al. (2016). The average of these slopes represent the 𝐸𝑎
𝑅 values.
(a) bare bars in TW (d) embedded bars in TW
(b) bare bars in SS (e) embedded bars in SS
y = 1.2225x + 0.1928R² = 0.9594
y = 1.2199x - 0.1669R² = 0.9793
y = 1.2175x - 0.5268R² = 0.9928
y = 1.2149x - 0.8866R² = 0.9994
0.0
1.0
2.0
3.0
4.0
5.0
2.0 2.5 3.0 3.5 4.0
ln(t
) (i
n d
ays
to a
ttai
n a
gi
ven
str
egn
th r
ete
nti
on
)
(1000/T)
60% SR
70% SR
80% SR
90% SR
y = 8.9727x - 16.593R² = 0.9956
y = 8.9135x - 18.817R² = 0.9995
y = 8.9906x - 22.496R² = 0.9934
y = 8.8768x - 25.072R² = 0.9999
0.0
3.0
6.0
9.0
12.0
15.0
2.0 2.5 3.0 3.5 4.0
ln(t
) (i
n d
ays
to a
ttai
n a
gi
ven
str
egn
th r
ete
nti
on
)
(1000/T)
60% SR
70% SR
80% SR
90% SR
y = 1.2645x + 1.0963R² = 0.9674
y = 1.2004x + 0.6415R² = 0.9775
y = 1.263x - 0.3481R² = 0.9858
y = 1.2312x - 0.9425R² = 0.9686
0.0
1.0
2.0
3.0
4.0
5.0
6.0
2.0 2.5 3.0 3.5 4.0
ln(t
) (i
n d
ays
to a
ttai
n a
gi
ven
str
egn
th r
ete
nti
on
)
(1000/T)
60% SR
70% SR
80% SR
90% SR
y = 4.6999x - 3.4274R² = 1
y = 4.6215x - 5.3191R² = 0.9797
y = 4.5149x - 7.3422R² = 0.9692
y = 4.557x - 9.635R² = 0.9989
0.0
3.0
6.0
9.0
12.0
15.0
2.0 2.5 3.0 3.5 4.0
ln(t
) (i
n d
ays
to a
ttai
n a
gi
ven
str
egn
th r
ete
nti
on
)
(1000/T)
60% SR
70% SR
80% SR
90% SR
(c) bare bars in AS (f) embedded bars in AS
Figure 10: Arrhenius plots for the service life of the bare bars and embedded GFRP bars in
concrete
The regression coefficient (𝐸𝑎
𝑅) were used to determine the relative time shift factor (𝑇𝑆𝐹) for
all conditioning cases with respect to two different temperatures [See Eq. (5)] as suggested by
Dejke (2001). In Eq. (5), 𝑡1 and 𝑡2 (days) are the time required to reach a certain strength
retention (𝑆𝑅), 𝑘1 and 𝑘2 are the degradation rates corresponding to 𝑡1 and 𝑡2, respectively,
𝑇𝑟𝑒𝑓 and 𝑇0 are the reference temperature and the exposure temperature (Kelvin), respectively.
Figure 12(a) illustrates the procedure on finding the TSF using Equation (5) based on the
corresponding exposure temperature. In this equation, the 𝐸𝑎
𝑅 values are determine from Figures
10 and 11 where Tref is taken as the room temperature of 23°C to evaluate the TSF for the target
temperature T0. For example, the TSF for an embedded GFRP bar in concrete immersed in TW
at 60°C is 28.64 as shown in Figure 12(a). Following this procedure, the corresponding TSF
values for all exposure conditions and temperatures considered in this study are tabulated in
Table 5. On the other hand, the TSF for an exposed temperature of 30°C shown in Figure 12(b).
A temperature of 30oC is considered as this is the average annual temperature in Australia.
Table 6 shows all values for embedded bars in concrete at 30°C with respect to 23°C. The main
purpose of 𝑇𝑆𝐹 is to transform the time taken in accelerated tests at a known exposure condition
and temperature, and correlate this with the actual service life of GFRP bars inside concrete.
In this study, the reference temperature was chosen at the room temperature condition of 23°C.
Accordingly, 𝐸𝑎
𝑅 and 𝑇𝑆𝐹 values for all conditioning cases are tabulated in Table 5. As can be
noticed from Table 5, the 𝐸𝑎
𝑅 values for embedded GFRP bars in concrete is much higher
compared to bare bars indicating that the required activation energy to cause the degradation
for the bars inside concrete is higher. This also means that the cement embedded bars have a
lower degradation rate compared to the GFRP bars which were directly immersed into the
solutions, which explains the benefit of the surrounding concrete cover in extending the
durability of the GFRP bars in the actual structures.
𝑇𝑆𝐹 =𝑡1
𝑡2=
𝑘2
𝑘1=
𝐴×𝑒(−𝐸𝑎𝑅𝑇0
)
𝐴×𝑒(
−𝐸𝑎𝑅𝑇𝑟𝑒𝑓
)= 𝑒
(−𝐸𝑎𝑅
)(1
𝑇0−
1
𝑇𝑟𝑒𝑓) Equation 5
y = 2.2104x - 3.3029R² = 0.9933
y = 2.174x - 3.5949R² = 0.9809
y = 2.2749x - 4.4127R² = 0.997
y = 2.3421x - 5.2947R² = 0.9521
0.0
1.0
2.0
3.0
4.0
5.0
2.0 2.5 3.0 3.5 4.0
ln(t
) (i
n d
ays
to a
ttai
n a
gi
ven
str
egn
th r
ete
nti
on
)
(1000/T)
60% SR
70% SR
80% SR
90% SR
y = 4.7851x - 7.2447R² = 0.9725
y = 4.583x - 8.6427R² = 0.991
y = 4.5986x - 10.681R² = 0.981
y = 4.6003x - 12.339R² = 0.9191
0.0
2.0
4.0
6.0
8.0
10.0
2.0 2.5 3.0 3.5 4.0
ln(t
) (i
n d
ays
to a
ttai
n a
gi
ven
str
egn
th r
ete
nti
on
)
(1000/T)
60% SR
70% SR
80% SR
90% SR
(a) TW at 60°C (b) Different exposure at 30°C
Figure 12: Illustration of evaluating the TSF for embedded GFRP bar in concrete
Table 5: 𝐸𝑎
𝑅 and 𝑇𝑆𝐹 values for all conditioning cases
Bar case Accelerated
agent
𝑬𝒂
𝑹
𝑻𝑺𝑭
23 °C 60 °C 80 °C
Bare GFRP
bars
TW 1217 1.00 1.58 1.94
SS 1240 1.00 1.59 1.97
AS 2250 1.00 2.33 3.41
Embedded
GFRP bars
TW 8946 1.00 28.64 131.06
SS 4599 1.00 5.61 12.26
AS 4642 1.00 5.70 12.55
Predicting the long-term behaviour and constructing master curves
A number of methods were suggested to predict the long-term behaviour of the GFRP bars
following accelerated tests (Bank, et al., 2003, Dejke, 2001). However, Chen et al. (2006) and
Ali et al. (2019) suggested that the better way of getting a more precise and accurate prediction
is to develop a master curve containing a plot for all the data used for analysis. This master
curve consist of the time required to reach a specific 𝑆𝑅 corresponded to its 𝑇𝑆𝐹 considering
the effect of temperature. Accordingly, Figures 13 and 14 show the master curves at a
temperature of 23oC for the bare and cement embedded bars at different conditioning
environment implemented in this study. These curves show the SR in % in the y-axis against
time, t in days in the x-axis. It can be observed that master curves can be expressed by a
logarithmic equation with 𝑅2 of at least 0.94. This means that the proposed 𝑆𝑅 versus time
model by (Bank, et al., 2003) is valid and resulted in a good prediction for the long-term
behaviour following Eq. (6) in its general form, where 𝑎 and 𝑏 are a regression constants.
𝑆𝑅 = 𝑎 ln(𝑡) + 𝑏 Equation 6
0.0
20.0
40.0
60.0
80.0
0 20 40 60 80
TSF
Temperature, T (°C)
28.64
0.0
1.0
2.0
3.0
4.0
20 30 40
TSF
Temperature, T (°C)
TW
SS
AS
1.44
2.01
(23°C)
(a) bare bars in TW (d) embedded bars in TW
(b) bare bars in SS (e) embedded bars in SS
(b) bare bars in AS (f) embedded bars in AS
Figure 13: Master curves for bare bars and embedded bars in concrete at 23 C°
Equivalent service life for GFRP bars in concrete environment
The strength retention measured for the bare GFRP bars exposed in the different solutions at
different temperature and exposure time was correlated to the strength retention of the cement
embedded bars to determine the equivalent service life for GFRP bars in the concrete
environment. As an example, the prediction of the service life of the GFRP bars in the concrete
environment was performed at a mean annual temperature of 30oC, which is the average annual
temperature in Australia. This prediction was made for ILSS strength retention as a function of
service life to a maximum of 100 years as bridge infrastructures in Australia are designed to be
in service for this length of time (Austroads, 2016). Accordingly, master curves of the SR of
the GFRP bars in concrete structures during its service life at 30°C exposed to different
SR = -26.99ln(t) + 176.69R² = 0.992
0
20
40
60
80
100
0 20 40 60 80 100
SR (
%)
Time , t (days)
SR = -3.55ln(x) + 108.53R² = 0.968
0
20
40
60
80
100
0 200000 400000 600000 800000
SR (
%)
Time , t (days)
SR = -22.98ln(x) + 174.98R² = 0.961
0
20
40
60
80
100
0 50 100 150 200
SR (
%)
Time , t (days)
SR = -3.725ln(x) + 109.01R² = 0.993
0
20
40
60
80
100
0 200000 400000 600000 800000
SR (
%)
Time , t (days)
SR = -23.21ln(x) + 157.35R² = 0.947
0
20
40
60
80
100
0 20 40 60 80 100
SR (
%)
Time , t (days)
SR = -4.569ln(x) + 102.06R² = 0.995
0
20
40
60
80
100
0 5000 10000 15000 20000
SR (
%)
Time , t (days)
exposure environments (TW, SS, and AS) were created (see Figure 15). These master curves
were constructed following the procedures described in developing Figures 15 and 16 but
modified using the corresponding 𝑇𝑆𝐹 listed in Table 6. As can be seen from the graphs, the
GFRP bars exposed in TW and SS can retain 60% of their ILSS up to 900 years and 1050 years,
respectively while GFRP bars in AS can retain 60% of its ILSS after 20 years in service.
Table 6 - 𝑇𝑆𝐹 values for embedded bars in concrete at 30°C with respect to 23°C
TW SS AS
𝑇𝑆𝐹 2.01 1.43 1.44
(a) TW (b) SS
(c) AS
Figure 15: Master curves for GFRP bars exposed at different environments at 30°C
From the master curves in Figure 15, the SR of the GFRP bars in concrete structures exposed
to different environment at an annual average temperature of 30oC up to 100 years was
established as shown in Figure 16. As highlighted in previous sections, the service life of the
GFRP bars in actual concrete structures was predicted from the accelerated aging test results
by correlating the strength retention of the bare GFRP bars to that of the cement embedded
bars. As a first step, the 𝑆𝑅 value for the bare bars was calculated using the curve fitting of the
master curves in Figure 12 and the corresponding 𝑇𝑆𝐹 value was calculated using Eq. (3).
Next, the master curve in Figure 13 for cement embedded bars was developed and modified
using the corresponding 𝑇𝑆𝐹. If the annual average temperature is at 30°C, then the master
curve in Figure 15 can be adopted directly. For example, the expected 𝑆𝑅 of a bare GFRP bar
submerged in SS and exposed to 60°C for 56 days using the master curve in Figure 14b is:
𝑆𝑅 = −22.98 ln(𝑡 × 𝑇𝑆𝐹) + 174.98 , where 𝑇𝑆𝐹 is 1.59 in Table 5.
SR = -3.571ln(x) + 85.035R² = 0.968
0
20
40
60
80
100
0 300 600 900 1200
SR (
%)
Time , t (years)
SR = -3.725ln(x) + 85.697R² = 0.993
0
20
40
60
80
100
0 300 600 900 1200SR
(%
)
Time , t (years)
SR = -4.569ln(x) + 73.452R² = 0.995
0
20
40
60
80
100
0 6 12 18 24
SR (
%)
Time , t (years)
𝑆𝑅 = −22.98 ln(56 × 1.59) + 174.98 = 71.82%
The equivalent service life for these GFRP bars in concrete structures exposed in alkaline
environment at an annual average temperature of 30°C using the master curve in Figure 14c,
i.e. (𝑆𝑅 = −4.569 ln(𝑡) + 73.452) where SR is 71.82% is 𝑡 = 1.429 years.
Another example, a bare GFRP bar submerged in TW for 30 days and exposed to 80°C, the
expected 𝑆𝑅 for this bar following the curve fitting in Figure 14a is:
𝑆𝑅 = −26.99 ln(𝑡 × 𝑇𝑆𝐹) + 176.69 , where 𝑇𝑆𝐹 is 1.94 in Table 5.
𝑆𝑅 = −26.99 ln(30 × 1.94) + 176.69 = 67.01%
The equivalent service life of these in an actual concrete environment at an annual average
temperature of 30°C and exposed to sea water (SS) using Figure 14b (𝑆𝑅 = −3.725 ln(𝑡) +109.01) is 𝑡 = 150.9 years, which is more than the service life (100 years) of the GFRP bars
in concrete structures. An accelerated test for bare GFRP bars submerged in AS for 20 days
and exposed to a temperature of 70°C will have a 𝑆𝑅 of 63.67% following the master curve in
Figure 13c:
𝑆𝑅 = −23.21 ln(𝑡 × 𝑇𝑆𝐹) + 157.35 , where 𝑇𝑆𝐹 can be calculated using Eq. (3).
𝑇𝑆𝐹 = 𝑒(−2250)(1
273.15+70−
1
273.15+23) = 2.831
𝑆𝑅 = −23.21 ln(20 ∗ 2.831) + 157.35 = 63.67%
The equivalent service life of these bars in actual concrete environment at an average annual
temperature of 30°C in an alkaline environment using Figure 14c, i.e. (𝑆𝑅 = −4.569 ln(𝑡) +73.452) is 𝑡 = 8.51 years.
Figure 16: Service life of GFRP bars in concrete structures at 30°C
As a summary, the GFRP bars will retain 54% of its ILSS when exposed to alkaline
environment and nearly 68% for bars exposed in tapwater and seawater after 100 years of
service (Fig. 15). These results further shows that the GFRP bars will last longer in the concrete
environment than directly exposed to the simulated concrete environmental conditions. These
findings support the observations by Benmokrane et al. (2018) wherein they measured a
maximum 16% reduction ILSS in vinyl-ester GFRP bars extracted from the concrete bridge
barriers after 11 years in service. The slightly lower ILSS reduction from the field study than
this laboratory study can be due to the bars were well protected by concrete as the concrete
cover is at least 65 mm. Similarly, the pH value measured from the core samples is only 12.3,
0
20
40
60
80
100
0 20 40 60 80 100
SR (
%)
Time , t (years)
TW
SS
AS
which is lower than the pH of the cement used in this study. Nonetheless, these results clearly
showed that the natural conditions are generally less aggressive to GFRP bars aging due to
lower temperature and humidity conditions than constant elevated temperature and continuous
contact and complete saturation in the solution of GFRP bars in the accelerated exposure tests.
Finally, these findings confirm the conclusions by Wang et al. (2017) that the long-term
predictions for FRP bars directly placed in the simulated solutions are too conservative
compared with the field results wherein the bars are embedded in concrete.
CONCLUSIONS
This study comparatively evaluated the durability of GFRP bars in concrete and in simulated
concrete environment through the investigation of their interlaminar shear strength. It focuses
on the evaluation of the physical, mechanical and micro-structural properties of GFRP bars
under high moisture, saltwater and alkali environments. From the results of this work, the
following conclusions can be drawn:
The percentage water uptake and the apparent diffusivity of the GFRP bars were
strongly dependent on the type of solution and temperature, with the percentage water
absorption and apparent diffusivity higher for high than low temperature exposure. For
similar degree of temperature exposure, the GFRP bars conditioned in the alkaline
solution has the highest moisture uptake and apparent diffusivity rate followed by the
bars conditioned in tap water with the saline solution the least.
The interlaminar shear strength of the GFRP bars decreased as the exposure
temperature and duration increased with the ILSS of cement-embedded GFRP bars
were generally higher than that of the bare GFRP bars for similar immersion conditions.
The alkaline solution is more aggressive to GFRP bars affecting its interlaminar shear
strength than tapwater and saline solution. After 112 days conditioning at 60oC, the bare
GFRP bars retained exposed to this solution retained only 30% of its interlaminar shear
strength with the bars exposed to tapwater and saline solution retaining 41% and 54%,
respectively.
Direct immersion in solution deteriorates the interlaminar shear strength of GFRP bars
more severely than in the case of cement embedded bars. After 112 days conditioning
at 80oC, the cement embedded GFRP bars exposed in alkaline solution can retain 68%
of its interlaminar shear strength compared to only 23% for bare GFRP bars.
SEM showed that the fiber surface of the cement embedded GFRP bars had more resin
coverage than the bare GFRP bars. Likewise, more residual resin covers on the fibre
surface was observed for the GFRP bars exposed to tap water and saline solution than
in alkaline solution suggesting a better bonding between the fibre and vinyl-ester resin.
The FTIR spectra did not show any significant changes in the polymers chemical
structure except for the relatively higher intensity of the O-H stretching band for the
GFRP bars directly immersed in the alkaline solution. However, this higher intensity
was only observed at the bar surface indicating that the water absorption was only
concentrated in the thin resin rich area of the GFRP bars.
Based on the Arrhenius relation, the required activation energy to cause the degradation
for GFRP bars inside concrete is higher than the bare GFRP bars directly immersed into
different accelerated aging solutions, which explains the benefit of the surrounding
concrete cover in extending the durability of the GFRP bars in the actual structures.
Master curves and time shift factor to correlate the strength retention of the the
accelerated aging test using bare GFRP bars to the equivalent service life for GFRP
bars in the concrete environment were developed. Based on this correlation, the GFRP
bars in actual concrete structures will retain up to 54%, 68% and 68% of its ILSS after
100 years of service at an annual average temperature of 30oC when exposed to alkaline
environment, tapwater and saline solution, respectively.
The results from this work provided a good representation and comparison of the long-term
properties and durability performance of GFRP bars in simulated and actual concrete
environment. Furthermore, the short shear beam shear test gave a straightforward and reliable
indication of the resistance of the fiber–matrix of GFRP bars exposed in different
environmental conditions. However, a comparative study to relate the interface property of
GFRP bars using short-beam shear test to the longitudinal properties can lead to a simpler and
more practical assessment of the durability and long-term performance of GFRP bars in
concrete environment.
References:
ASTM D7705/D7705M – 12: Standard Test Method for Alkali Resistance of Fiber
Reinforced Polymer (FRP) Matrix Composite Bars used in Concrete Construction
ASTM D 1141-98 (2013): Standard Practice for the Preparation of Substitute Ocean Water
ASTM D4475 -02 (2016): Standard Test Method for Apparent Horizontal Shear Strength of
Pultruded Reinforced Plastic Rods by the Short-Beam Method.
ASTM D570 (2010): Water absorption of plastics. ASTM D4972-13, ASTM International,
West Conshohocken, Philadelphia, Pa 19103.
ASTM Standard ASTM D4972 (2013). Standard Test Method for pH of Soils. ASTM D4972-
13, ASTM International, West Conshohocken, Philadelphia, Pa 19103.
Adams, D. 2018. What is the most important type of mechanical test for composites?
Composites World 2018; 4(1), 10-11.
Aiello, M.A., Leone, M., Aniskevich, A.N., Starkova, O.A. (2018). Moisture effects on
elastic and viscoelastic properties of CFRP rebars and vinylester binder. ASCE Journal
of Composites for Construction, 18, 686-691.
Ali, A. H., Mohamed, H. M., Benmokrane, B., ElSafty, A., and Chaallal, O. (2019).
"Durability performance and long-term prediction models of sand-coated basalt FRP
bars." Composites Part B: Engineering, 157, 248-258.
Almusallam TH, Al-Salloum YA, Alsayed SH, El-Gamal S, and Aqel M. Tensile properties
of glass fiber-reinforced polymer bars embedded in concrete under severe laboratory
and field environmental conditions. Journal of Composite Materials 2012; 47(4), 393-
407.
Ashrafi, H, Bazli, M, Oskouei AV, Bazli, L. (2018). Effect of Sequential Exposure to UV
Radiation and Water Vapor Condensation and Extreme Temperatures on the
Mechanical Properties of GFRP bars. ASCE Journal of Composites for Construction,
22(1), 04017047-1-17.
Austroads (2016).Realising 100-years Bridge Design Life in an Aggressive Environment:
Review of the Literature. Austroads Technical Report AP-T313-16.
Bank, L. C., Gentry R.T., Thompson, B.P., Russell, S.J. (2003). A Model Specification for
FRP Composites for Civil Engineering Structures. Construction and Building
Materials, 17(6-7), 405-437.
Belarbi A. and Wang, H. (2012). “Durability of FRP bars embedded in fiber-reinforced
concrete.” ASCE Journal of Composites for Construction, 16(4), 371-380.
Benmokrane B, Ali A, Mohamed H, Robert M, and ElSafty A. (2017a). “Durability
Performance and Service Life of CFCC Tendons Exposed to Elevated Temperature and
Alkaline Environment”. ASCE Journal of Composites for Construction; 20(1),
04015043-1:13.
Benmokrane B, Manalo A, Bouhet J-C, Mohamed K, and Robert M. (2017b). “Effects of
diameter on the durability of glass-fiber-reinforced-polymer (GFRP) bars conditioned
in alkaline solution.” ASCE Journal of Composites for Construction; 21(5), 04017040-
1:12.
Benmokrane B, Ali AH, Mohamed HM, ElSafty A, Manalo A. Laboratory assessment and
durability performance of vinyl-ester, polyester, and epoxy glass-FRP bars for concrete
structures. Composites Part B. 2017; 114: 163-174.
Benmokrane B, Nazair C, Loranger M, and Manalo A. (2018). “Field Durability of Vinyl-
Ester-Based GFRP Rebars in Concrete Bridge Barriers”. ASCE Journal of Bridge
Engineering, 23(12), 04018094-1-13.
Ceroni, F, Cosenza, E, Gaetano, M, Pecce, M. Durability issues of FRP bars in reinforced
concrete members. Cement and Concrete Composites 2006; 28, 857-868.
Chen, Y., Davalos, J. F., and Ray, I. (2006). "Durability prediction for GFRP reinforcing bars
using short-term data of accelerated aging tests." Journal of Composites for
Construction, 10(4), 279-286.
Chin, J.W., Aouadi, K., Haight, M.R., Hughes, W.L., and Nguyen, T. (2001). “Effects of
water, salt solution and simulated concrete pore solution on the properties of composite
matrix resins used in civil engineering applications”. Polymer Composites, 22(2), 282-
297.
Davalos, J.F., Chen, Y., and Ray, I. (2012). “Long-term durability prediction models for
GFRP bars in concrete environment”. Journal of Composite Materials, 46(16), 1899-
1914.
D’Antino T., Pisani, M.A., and Poggi, C. (2018). “Effect of the environment on the
performance of GFRP reinforcing bars”. Composites Part B, 141, 123-136.
Dejke, V. (2001). "Durability of FRP reinforcement in concrete: Literature review and
experiments." Chalmers University, Sweden.
Ferdous, W., Manalo, A., Khennane, A. and Kayali, O. (2015). “Geopolymer concrete-filled
pultruded composite beams - Concrete mix design and application”. Cement and
Concrete Composites, 58, 1-13.
Gooranorimi, O and Nanni, A. (2017). “GFRP Reinforcement in Concrete after 15 Years of
Service”. ASCE Journal of Composites for Construction, 21(5), 04017024-1-9.
Kamal A, Boulfiza M. Durability of GFRP rebars in simulated concrete solutions under
accelerated aging conditions. ASCE Journal of Composites for Construction 2011;
15(4), 473-481.
Karbhari, V.M., Murphy, K. and Zhang, S. (2002). “Effect of concrete based alkali solutions
on short-term durability of E-glass/vinyl ester composites”. Journal of Composite
Materials, 36(17), 2101-2121.
Karbhari, V.M. and Zhang, S. (2003). “E-glass/vinyl ester composites in aqueous
